The statements in this section merely provide background information is related to the present disclosure and may not constitute the prior art.
MPEG (Moving Picture Experts Group) and VCEG (Video Coding Experts Group) developed an improved and excellent video compression technology over the MPEG-4 Part 2 and H.263 standards. The new standard is called H.264/AVC (Advanced Video Coding) and was released simultaneously as MPEG-4 Part 10 AVC and ITU-T Recommendation H.264.
In the image compression techniques such as H.264/AVC, various prediction techniques are employed including an inter prediction and an intra prediction to obtain effective image compressions. Here, the intra prediction includes an intra_4×4 prediction, intra_16×16 prediction, intra_8×8 prediction and other various predictions, where the respective intra predictions include a plurality of prediction modes.
The image compression techniques such as H.264/AVC compresses each block in a plurality of prediction modes from which an optimal prediction mode is selected to perform a prediction. Therefore, the compression efficiency depends on a particular prediction mode selection as the optimal prediction mode to perform the prediction, and the optimal prediction mode selection requires block predictions in the entire prediction modes and followed by calculations of the encoding cost using a predetermined cost function to select the prediction mode with the lowest encoding cost.
Therefore, in order to determine the optimal prediction mode whether it is intra_4×4 prediction or intra_6×16 prediction, the cost should be calculated after trying every one of the prediction modes through the predictions. This makes an encoding process complex and lowers the encoding efficiency or compression efficiency.
Additionally, to correctly decode the image, information on the selected optimal prediction mode should be also encoded. In response to recent incremental requirements for higher resolution videos, the number of blocks to be encoded has been increased resulting in a greater quantity of bits required to encode the information on the optimal prediction mode with respect to each block, which degrades the compression efficiency.